Tap Tremolo Pre-Planning Questions

[transcendent12]

I've kind of been thinking about this for the last few years cause my brother has always been bugging me to make him a tremolo pedal (At the time I was an EE student).

I lack a bit of practical circuit design skills, still very green and have very little general knowledge when it comes to the basic circuits and characteristics =)

Anyhow, I realise there is already a wealth of information upon the web and this forum on existing tremolo's but I'd like to clarify a few things.

My tremolo is based upon a few things for initial prototype. I'm using a 32bit Arduino Due to digitally control and synthesize the tremolo waveforms using it's 12bit DAC.

I have seen a few classes of tremolo methods: 1) Optocoupled somehow 2)direct modulation of various transistors  3) Various types of Variable Gain Amplifiers.

Features of the my proposed tremolo: Tap tremolo (hardware interrupts into arduino being counted and averaged), BPM selector, memory functions, programmable waveforms and regular features such as depth etc, true bypass.

From what I understand, the opto coupled tremolo does not have a linear response between incident light and resistance. While this may make a pleasing sound, I'd rather just IQ record/synthesize an LDR circuit and then program them into the arduino. So for this I need a "bit perfect" linear Variable Gain Amplifier.

I haven't seen anyone else use this particular chip, so I'd be appreciative if anyone could say if it was a reasonable, or terrible, idea.
LT1251 http://cds.linear.com/docs/en/datasheet/12516fa.pdf

Naturally I have no idea of the acoustic characteristics of this device and it could make the guitar signal sound like trash after it's done with it. But I figured it might be worth a try. This is obviously a fairly simple solution. If anyone has any other ideas using well known opamps in a topology which allows linear voltage controlled gain, that would be great too.

So all in all, the whole box would have a whole lot of digitally controlled functions, a single amplifier IC and some power circuitry, possibly some gain signal compensation if the amp can't handle it.

[VidSicious]

Tldr but maybe you should look into building a vero version of tap tremolo. Here's a layout that's been posted recently

http://tagboardeffects.blogspot.co.uk/2015/10/twin-peaks-tap-tempo-tremolo.html?m=1

[transcendent12]

Tldr but maybe you should look into building a vero version of tap tremolo. Here's a layout that's been posted recently

http://tagboardeffects.blogspot.co.uk/2015/10/twin-peaks-tap-tempo-tremolo.html?m=1

Thanks, a lot of good information which is very enlightening. I'd still like play around with my own build for educational purposes too, but there are many good ideas there =)

[samhay]

Your planning on using the DAC to generate a control voltage, which is unusual. As you have probably realised, most [edit - DIGITAL] LFOs in the DIY realm use a PWM output to drive LEDs. This is simple, and was probably the only viable some years ago (pre Arduino, etc). The non-linearity between LED light output and LDR resistance (and also current/control voltage to LED light output) can be overcome by using a lookup table to generate the LFO values, but to my ear a sine wave LFO driving a LED/LDR sounds as good as any wave shape I have yet tried.

I have no experience with the LT1251/6, but it looks like an interesting chip. Have you established whether you can actually purchase one from somewhere in single units?

[PRR]

> opto coupled tremolo does not have a linear response between incident light and resistance.

Over two decades (perhaps 500K-5K) it is within 40%. Typical trem depths don't run much past 1 decade (if a signal is cut 10:1 quickly, it is "gone"). The ear's linearity isn't anywhere near 10%. The accuracy is good. (And if you have a CPU in the box, you can have a calibration table.) OTOH opto offers *complete* isolation between audio and digi-crap, which can be a major problem.

LT1251 cuts-to-Zero at the ends; LT1256 is linear to the extremes. For mild wobble, no different; in extreme wobble the abrupt cut-off of the '51 may call attention to itself. (And again, if you have a CPU the you can program-in/out any abrupt cut-off/not that you wish.)

Audio performance is likely excellent within obvious limits (output won't reach supply rails and input limit is a bit lower). Hiss seems fine and THD is off the chart. Input impedance may be fine without buffer? (Input bias current is marginal.) While rated to 36V, most specs are for +/-5V so there is hope it will be happy on 9V.

A 128-step Digital Pot (buffered) is probably all the resolution you need, and may be much smaller, simpler, less to wire.

[PRR]

> can actually purchase one..?

8 bux direct from Linear:
http://www.linear.com/purchase/LT1251
http://www.linear.com/purchase/LT1256

8 bux on eBay from Hong Kong:
http://www.ebay.com/itm/LT1251CN-LT1251-DC-Gain-Controlled-Amplifier-IC-/270791848805
http://www.ebay.com/itm/LT1256CS-40MHz-Video-Fader-and-DC-Gain-Controlled-Amplifier-IC-/280910758948

Check package type (DIP or SMD) and shipping costs.

Also Linear's shop-cart defaults to Quan 2, which is a wise idea, but if you want ONE! it may not cooperate?

[PRR]

That video-fader was setting off bells in my head.

Turn to page 85 of this book:
The Art and Science of Analog Circuit Design, Jim Williams, 1998

William H. Gross designed that chip, and when Jim asked him for an essay, Gross wrote a VERY introspective account of the winding path he followed to develop that chip.

You can peek at Gross' essay via Amazon "Look inside" feature:
http://www.amazon.com/Science-Analog-Circuit-Design-Engineers/dp/0750670622
Scroll to page 85 without reading too many other pages (there is a limit what they let you peep for free).

While even used copies of tAaSoACD ask high prices, to a design student or apprentice this book is VERY worth the price.

> I was an EE student

If you have not spent thousands of hours in the attic (pg 270) staring at hallucinated circuits, Gross' essay may be TMI.

> can actually purchase one

Good-good question. From the essay, and the data-sheet properties, we know this chip is 18 years old. This is "forever" in chip-world. The fader had wide application in analog video, but increasingly video processing is done in a CPU. Also the 2-second Google search did not pop-up the Usual Suspects like DigiKey or Mouser, meaning they are not listing it with Google Ads. While available today, its sun is probably setting. So Linear's shopping-cart pushing "2" at you is a good idea twice: you might blow-up the first one; if not, Brother will come to you in 2025 or 2035 with a blown chip and expect you to pull a spare out of your ear. Meanwhile a lot of old video-desks nibble at the last batch. It may be tough to source a replacement many years down the road.

I had this problem. Good 1977 chip but I abused it enough to kill a chip in 8 years, and again 8 years later. By then it was NOT readily available. I just managed to get 2 more from the Last-Call! batch, which covered me until I retired (no longer my problem).

[bluebunny]

[ASIDE]
What an interesting book, Paul.  I'm not sure I really need it (though at £25, it's tempting).  But the first chapter - "The Importance of Fixing" - is a fascinating read (and it's complete in the amazon.co.uk "Look Inside").
[/ASIDE]

[transcendent12]

That video-fader was setting off bells in my head.

Turn to page 85 of this book:
The Art and Science of Analog Circuit Design, Jim Williams, 1998

William H. Gross designed that chip, and when Jim asked him for an essay, Gross wrote a VERY introspective account of the winding path he followed to develop that chip.

You can peek at Gross' essay via Amazon "Look inside" feature:
http://www.amazon.com/Science-Analog-Circuit-Design-Engineers/dp/0750670622
Scroll to page 85 without reading too many other pages (there is a limit what they let you peep for free).

While even used copies of tAaSoACD ask high prices, to a design student or apprentice this book is VERY worth the price.

> I was an EE student

If you have not spent thousands of hours in the attic (pg 270) staring at hallucinated circuits, Gross' essay may be TMI.

> can actually purchase one

Good-good question. From the essay, and the data-sheet properties, we know this chip is 18 years old. This is "forever" in chip-world. The fader had wide application in analog video, but increasingly video processing is done in a CPU. Also the 2-second Google search did not pop-up the Usual Suspects like DigiKey or Mouser, meaning they are not listing it with Google Ads. While available today, its sun is probably setting. So Linear's shopping-cart pushing "2" at you is a good idea twice: you might blow-up the first one; if not, Brother will come to you in 2025 or 2035 with a blown chip and expect you to pull a spare out of your ear. Meanwhile a lot of old video-desks nibble at the last batch. It may be tough to source a replacement many years down the road.

I had this problem. Good 1977 chip but I abused it enough to kill a chip in 8 years, and again 8 years later. By then it was NOT readily available. I just managed to get 2 more from the Last-Call! batch, which covered me until I retired (no longer my problem).

wow! fascinating, I had no idea it was that old. Even more surprising is a chapter of a book being written about it.

[bluebunny]

Look what you made me do, Paul!   



Just spent quite a lot of the morning dipping in and out of the Amazon preview.  What a great collection of essays!

[PRR]

DigiKey has the LT1256 at $11, but only 34 of the PDIP in stock (over 500 of the SOIC).

The LT1251 (hard cut-off) version they only have 7 SOIC.

Other parts are listed but minimum order is thousands!

Mouser never heard of it.

Jameco.com (former king of old chips) only has LT1256 in SOIC at $15.

I found LT1256 on Conrad *Bosnia* at a price which looks like $17USD.

[transcendent12]

Ordered a couple of samples for now =D
Had a thorough look at other VCA's and none of them seem to fit the bill as well as the LT1256. Mostly linear in dB control and higher gain ranges not going much below 0 to -30 db. Which isn't quite quite as fit for purpose. Also less control over the absolute voltage range as reference.
The PWM controlled optocoupler does seem pretty good, now that I have looked into it. But I kinda got an itch to be different for the sake of it XD. The extra isolation does sound nice though..certainly an overall more efficient topology